Many of today’s DSP applications are subject to real-time constraints. Many embedded DSP applications will eventually grow to a point where they are stressing the available CPU, memory or power resources. Understanding the workings of the DSP architecture, compiler and application algorithms can speed up applications, sometimes by an order of magnitude. This chapter will summarize some of the techniques that can improve the performance of your code in terms of cycle count, memory use and power consumption.
Optimization is a procedure that seeks to maximize or minimize one or more performance indices. These indices include:
• Throughput (execution speed)
• Memory usage
• I/O bandwidth ...